원문정보
초록
영어
Task scheduling under a grid environment is an important research area, on which much
attention has been paid. However, either in the meta-task scheduling problems or DAG
(Direct Acyclic Graph) scheduling problems, it is usually assumed that tasks are submitted to
dedicated hosts and that these tasks are processed in FIFO (First In First Out) order. This is
not practical in a grid, in which a host may be shared between grid users and its owner and
local tasks, which belong to resource owners, may compete with grid tasks for the hosts.
EBGSA (Estimation Based Grid Scheduling Approach) is proposed, which allows for the
simultaneous processing of grid tasks and local tasks. In EBGSA we use history information
about the execution of tasks to estimate the performance of non-dedicated hosts. Two
heuristic scheduling algorithms, MCT (Minimum Completion Time) and Min-min, are
selected to perform the simulation experiment. Both experiments obtain a smaller makespan,
proving EBGSA feasible for grid task scheduling.
목차
1. Introduction
2. Problem Definition
2.1. Performance Metrics
2.2. Problems with Existing Algorithms
3. EBGSA
3.1. Applying EBGSA to MCT and Min-min
3.2. Estimating the Time Variable
4. Simulation
4.1. Simulation Environment
4.2. Simulation Procedures
4.3. The Result and Discussion
5. Conclusion
References